X-MADAM-RAG：诊断和处理检索增强生成中的中英文证据冲突
X-MADAM-RAG: Diagnosing and Handling Chinese-English Evidence Conflict in Retrieval-Augmented Generation
AI总结：
提出X-MADAM-RAG管道，通过分解证据处理步骤（候选提取、可见证据修复、确定性分组和冲突感知聚合）解决RAG中中英文证据冲突问题，在受控基准上取得高准确率，但发现文档级提取是主要瓶颈。
摘要：
检索增强生成（RAG）系统可能接收到不仅噪声大而且相互矛盾的证据。这个问题在多语言环境中尤为突出，因为检索到的中文和英文证据可能支持不相容的答案候选。我们通过X-RAMDocs-ZHEN（一个从RAMDocs衍生的受控中英文基准）研究此问题，用于诊断RAG中的证据冲突。该基准包含300个示例，涵盖六种平衡条件，包括单语言支持、双语一致、反向冲突方向以及带可选噪声的冲突。我们进一步研究了X-MADAM-RAG，一个可解释的管道，将证据处理分解为每个文档的候选提取、可见证据修复、确定性候选分组和冲突感知聚合。在原始受控基准上使用Qwen2.5-7B-Instruct，X-MADAM-RAG达到了0.9667的严格准确率和0.9767的冲突感知成功率，优于证据归一化的单次调用基线。然而，一个零调用的纯规则提取器在同一基准上达到了1.0000，揭示了强模板规律性。为了探究这一局限性，我们构建了一个确定性自然化压力测试，移除了显式答案模板但保留了候选字符串。在其100样本子集上，纯规则提取器降至0.0000，但X-MADAM-RAG也降至0.3000严格准确率，低于朴素基线和证据归一化基线。特权Oracle保持完美，表明文档级提取是主要瓶颈。这些发现将X-RAMDocs-ZHEN和X-MADAM-RAG定位为受控证据冲突的诊断工具，而非通用幻觉检测或对自然检索鲁棒性的证据。
英文摘要：
Retrieval-augmented generation (RAG) systems may receive evidence that is not merely noisy but mutually contradictory. This issue becomes particularly salient in multilingual settings, where retrieved Chinese and English evidence may support incompatible answer candidates. We study this problem through X-RAMDocs-ZHEN, a controlled Chinese-English benchmark derived from RAMDocs for diagnosing evidence conflict in RAG. The benchmark contains 300 examples across six balanced conditions, including monolingual support, bilingual agreement, reversed conflict directions, and conflict with optional noise. We further examine X-MADAM-RAG, an interpretable pipeline that decomposes evidence handling into per-document candidate extraction, visible-evidence repair, deterministic candidate grouping, and conflict-aware aggregation. On the original controlled benchmark with Qwen2.5-7B-Instruct, X-MADAM-RAG achieves 0.9667 strict accuracy and 0.9767 conflict-aware success, outperforming an evidence-normalized single-call baseline. However, a zero-call rule-only extractor reaches 1.0000 on the same benchmark, revealing strong template regularity. To probe this limitation, we construct a deterministic naturalized stress test that removes explicit answer templates while preserving candidate strings. On its 100-sample subset, rule-only extraction falls to 0.0000, but X-MADAM-RAG also drops to 0.3000 strict accuracy, below both naive and evidence-normalized baselines. A privileged oracle remains perfect, indicating that document-level extraction is the main bottleneck. These findings position X-RAMDocs-ZHEN and X-MADAM-RAG as diagnostic tools for controlled evidence conflict rather than as evidence of general hallucination detection or robustness to natural retrieval.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
